# Handtrack

A library for prototyping realtime hand detection (bounding box), directly in the browser.

Repository: https://github.com/victordibia/handtrack.js
Canonical: https://ross.abutalabs.com/products/handtrack
Homepage: https://victordibia.com/handtrack.js/
Language: JavaScript
License: MIT
License Family: permissive
Topics: javascript, javascript-library, handtracking, neural-networks, artificial-intelligence
Last push: 2022-07-20T05:33:23+00:00
Link (homepage): https://victordibia.com/handtrack.js/

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 2767, "days_push": 1505, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2930, forks 258 (observed 2026-08-28T04:07:30.600199+00:00)

## What it is
Handtrack.js is a JavaScript library for prototyping realtime hand detection directly in the browser, framing handtracking as an object detection problem solved with a trained convolutional neural network. It predicts bounding boxes for hands (and hand poses like open, closed, pinch, point) from webcam or image input without requiring users to train their own models.

## Use cases
- detect hands in webcam video in the browser
- build gesture-based game controls
- track hand poses like pinch, point, open, closed
- detect face touching to minimize covid spread
- create air guitar or motion-controlled demos
- prototype hand tracking without training a model

## When to choose
- you need quick, in-browser hand detection with no model training
- you want a lightweight prototype using webcam input
- you need bounding-box hand tracking plus a few hand pose classes

## When to avoid
- you need production-grade accuracy or fine-grained skeletal hand landmarks
- you need server-side or native mobile performance
- you need the full 21-joint hand keypoint tracking other libraries provide

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, computer-vision, image-processing
- domain: computer-vision, machine-learning, web-development, artificial-intelligence
- platform: browser
- tags: hand-detection, handtracking, object-detection, tensorflow-js, webcam, realtime, bounding-box, gesture-recognition, javascript, web

## Member repositories
- victordibia/handtrack.js (sdk) score 23
- victordibia/handtracking (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:30.600199+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T07:33:25.251348+00:00, confidence not recorded.
  - readme: https://github.com/victordibia/handtrack.js (fetched 2026-08-28T04:07:30.600199+00:00, sha b9eb5d204b17)
  - homepage: https://victordibia.com/handtrack.js/ (fetched 2026-08-29T09:48:59.709664+00:00, sha 1d8d6dc6206d)
- Data as of 2026-08-30T08:39:29.467469+00:00.